A new application of information hiding in visual surveillance is presented in this research. The proposed framework is composed of three main components. The first component is a moving object detection mechanism, in which a static camera with a broader view of the scene is responsible for determining the objects to be tracked. The second component utilizes the information of estimated position/size of detected object to control the Pan-Tilt-Zoom (PTZ) camera for capturing sufficiently large regions of interest (ROI) in the targets. The third component is an information hiding scheme, which embeds the ROI acquired from the PTZ camera and the related labels into the video recorded by the static camera. The process of retrieving the ROI can thus be facilitated. Experimental results will demonstrate the feasibility of this interesting framework.